A STUDY OF THE PHENOMENON
OF TOTTING FILMS

BY
KENNETH LYLE WARREN

A THESIS
Presented to the Graduate School of Michigan State College
of Agriculture and Applied Science in Partial Fulfillment
of Requirements for the Degree of Doctor of Philosophy

Physice Department
East Lansing, Michigan
1935

ProQuest Number: 10008457

All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.

uest
ProQuest 10008457
Published by ProQuest LLC (2016). Copyright of the Dissertation is held by the Author.
All rights reserved.
This work is protected against unauthorized copying under Title 17, United States Code
Microform Edition © ProQuest LLC.
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 4 8 1 0 6 - 1346

Ac know 1 e dge me nt

I wish to express my sincere appreciation and
gratitude to Dr. C.W. Chamberlain of the Physics
Department of Michigan State College, whose helpful
criticisms and suggestions were a constant source of
inspiration in the experimental work as well as in
the preparation of this thesis.

Table of Contents
Article
Introduction

1

Wringing Film

1

Adsorbed Film

3

Wetting Film

4

Liquids which Produce Wetting Films

4

Liquids which Produce No Wetting Films

5

Wetting Films from Mixtures of Liquids not
Producing Wetting Films
Saturated and Unsaturated Hydrocarbons

6

8

Low Temperature Combustion

11

Early Observers of Wetting Films

14

Surface Tension

16

Effect of Saturated Vapor above Gasoline Surracea
Measurements of Pseudo Viscosity

19

Method of Measurement

20

Thickness of Wetting Films

24

Flow of Liquid in Wetting Films

27

The Effect of Age and Temperature on the
Surface Tension of Naphtha

29

Evaporation at Curved Surfaces

29

Summary

37

References

39

Bibliography

40

A Study of the Phenomenon
of Wetting Films

The wetting method 01 measuring the depth of a liquid
in a container is extremely inaccurate;

this is especially

true in the case of gasoline, naphtha and alcohol.

In

attempting to solve the problem of measuring accurately the
depth of gasoline in an underground tank a

problem which is

developing serious labor difficulties and enormous economic
loss, a very curious phenomenon was observed.

Gasoline was

found to wet a vertical glass plate to a height much greater
then could be accounted for by the phenomenon of capillarity.
This phenomenon has been termed the wetting film, to
distinguish it from the wringing film and the adsorbed film.
Wringing Film.

Rolt

(1) in speaking of the "Wringing”

of Block Gauges says, "A most valuable feature of these
gauges is the fact that it is possible to "wring” two or
more gauges together, by first cleaning them and then b r i n g ­
ing their faces into contact with a combined sliding and
twisting motion.

The adherence is due almost entirely to

the presence of a liquid film between the faces.-

It is

interesting to note that this minimum th?ckness found for

2

wringing films is roughly twice the accepted value of the
radius of molecular attraction- a relationship which offers
explanations for most of the properties associated with the
phenomenon of wringing.”

The radius of molecular attraction

was measured by Chamberlain (2) in 1910*
Quincke

(3), in 1869, made an attempt to measure the

Radius of Molecular Attraction, hereafter designated as R,
and concluded that it was greater than 5 x 10"® cm*

It was

generally agreed that a liquj-d film could not exist less
than 3R thick*

Reinhold and Rucker,

in 1881 (4), 83 (5),

and 93 (6 ) measured liquid films 1.2 x 10”^ cm. thick#
Johonott and Chamberlain (7), In 1899, measured films
0.6

x 1 0 "® cm. thick, and observed the splitting of the

Reinhold and Rucker film, and the reverse phenomenon of
doubling.

These results made it impossible to longer accept

Quincke*s determination of R#
As the value of R is far beyond the limit of the best
compound microscope it was necessary to employ a new type
of measuring instrument.

Chamberlain designed the Compound

Interferometer at Denison University and with the assistance
of Ernst Keil, instrument maker of the Department of Physics
in Barney Laboratory,

constructed the instrument shown below*

3

Til© optical.parts, possessing an accuracy^ greater .th^n on©
twentieth or a light wav©, were figured by O.L.Petitdidier*
Th© value of fl was found to b e 1 ♦5 x 10”^ciri*, or on© and
a half millionths of a millimeter*

Films one R thick are

twenty times smaller than c a n be observed by the best
compound microscope and, in the .case o f water, is
approximately five molecules thick*
Adsorbed F i l m *

An adsorbed film is a layer of a liquid

or gas which has condensed on a solid surface •
manifestation of the force of adhesion*

It is a

4

Wetting; Film*

A wetting film is that film which spreads

over a solid or liquid surface making contact at an angle
with the liquid surface*
Liquids which Produce Wetting Films *

While the present

investigation was undertaken to determine the cause of the
large wetting film or gasoline, and the varying character of
this film, many liquids were examined to determine which ones
possess the ability to form wetting films, and possibly throw
light upon the cause of the phenomenon.
Ethyl alcohol (CgH^OH) has a wetting film which creeps
very rapidly up the surfaces of silver and glass.
ments on alcohol were not taken for two reasons:

Measure­
first, because

of the work of Langmuir and second, because it seemed advisable
to confine the experimental procedure to hydrocarbons.
Results of this investigation show why alcohol creeps up
the sides of a glass container.

Langmuir (8 ) gives an

explanation of the spreading of organic molecules, having
the -COOH, -CO, and -OH radicals, over liquid surfaces.
While it is true that alcohol creeps up a vertical glass
plate the glass plate must have on it a wringing film, as
has been shown by Chamberlain*
Ether was found, to creep producing a very thin
wetting film.

5

Gasoline forms a wetting film in rather a violent
manner*

It was impossible to account for the wetting film

by capillarity*
two methods;

In general new films can be produced by

first, disturbing the surface, producing a

small wavelet;

second, by introducing a small amount of

air near the surface of the gasoline.

Different samples of

gasoline were tried as well as treated and untreated gas­
oline and In each case a wetting film was formed.
A sample of ordinary cleaning naphtha was found to
produce a wetting film, but not so great as that or gasoline.
Turpentine, which has as its main constituent pinene,
was found to act very similar to naphtha, producing an
appreciable wetting film.
Liquids which Produce no Wetting Films *

Acetone and

benzene form no wetting films.
Castor oil (C^ 7 Hgg(OH)COO) 3 C 3 H 5 does not creep up the
sides of a vertical container but it does spread over the
surface of water according to L angmuir1s theory, if placed
on the water in small quantities.
Distillate,

a product formed during the distillation

of crude oil, does not creep.
Fuel oil grade HA M which contains hydrocarbons of high
molecular weight was found not to creep*
Treated and untreated kerosenes do not p r o d u c e wetting

6

films*

By a treated kerosene or gasoline is meant one which

has been treated with sodium plumbite

(Na^PbC^)

to remove

the sulfur.
The fact that kerosene does not creep is contrary to
statements made by many authors.

This can be explained by

the conclusions reached in this investigation.

The old

kerosene, because of the commerical demand, was distilled
between a wide range of temperature.

The demand for modern

gasolines has increased to such an extent that the range over
which gasoline is distilled is increased; making it necessary
to decrease the limits for the distillation of kerosene.
Hydrogen peroxide and dilute acetic acid (CH 3 COOH)

do

not creep.
Xylene gave results which were very similar to benzene
showing no wetting film.
Cenco pump oil, paraffin oil and a medium grade of motor
oil were found to produce no wetting film.
Wetting Films from Mixtures of Liquids not Producing
Wetting F i l m s ♦

In mixing different liquids in an attempt to

stop the formation of the wetting film and also to see if the
effect could be increased,
observed.

a very interesting phenomenon was

It was found that certain hydrocarbons which in

themselves would not produce a wetting film did have a w e t ­
ting film when mixed.

In Table I below are the percentages

of liquids used, when the first observable wetting film was

7

noted*
Table X
Cenco Pump oil---

50$,

Xylene'

50$

Paraffin o il------

•33 l/3$,

Fuel oil Grade "A”

•50$,

Xylene--

Kerosene -----------

•50$,

Xylene----- —

Xylene-------- -----

•50$,

Benzene---- -----------

Cenco Pump oil----

■50$,

Benzene--------------------- -50$

Xylene

■66
- 2/3$
~ ^

---------------50$
---

----50$
50$

Paraffin oil---------33 l/3$,

B e n z e n e - - ---------

Fuel oil Grade "A"------ 50$,

Benzene---------------------- 50$

Kerosene----------------- 50$,

Benzene---------------------- 50$

Benzene-------------------50$,

Ether---------------------- —

Xylene--------------------50$,

Ether-

-66

2/3$

-------------------50$

Paraffin oil 12|r$, Cenco pump oil 12-J-$, Ether 12|r$, and
Benzene 62-J-$*
In connection with the mixing of benzene and xylene
with ether it must be r emembered that It was previously
stated that ether had a very thin wetting film*
The wetting film in some of the above cases is slight
but it was found that if the proportion of xylene or benzene
was Increased the rapidity with which it rises increases
as the amounts of xylene and benzene are increased*
It should also be noted that all mixtures containing

8

benzene and xylene contain equal amounts of each, component.
Saturated and Unsaturated H y drocarbons.

The cracking

process produces unsaturated hydrocarbons in the gasoline.
To investigate their effect in producing the wetting film
the unsaturated hydrocarbons were removed.

This was done

by two methods.
First method.

Fifty cc. of concentrated sulfuric acid

(HgSO^) were mixed with 500 cc. of gasoline and the resulting
mixture placed in a mechanical shaker for two and a half
hours.

The mixture was then placed in a separator funnel

and the oxidized gasoline and carbon compounds were poured
out the bottom of the funnel.

To the remaining gasoline

another 50 c c . of concentrated sulfuric acid was added and
the mixture again placed in the mechanical shaker for two
hours.

Again the oxidized gasoline and carbon compounds

were separated out by the separator funnel.
With a small sample of gasoline the bromine test for
unsaturated hydrocarbons was made.

The testing liquid was

a solution of water saturated with bromine.

The disappear­

ance of the bromine color indicated that all of the unsatu­
rated hydrocarbons had not been saturated by the sulfuric
acid.

It was now necessary to add another 50 c c . of concen­

trated sulfuric acid to the gasoline and shake as befoie until

9

til© gasoline was completely saturated*

To tills saturated

gasoline a 10$ solution of sodium carbonate (Na 2 C03 ) (10
grams of UagC 0 3 ant^

cc •

water) was added*

This m i x ­

ture was placed in the mechanical shaker for fifteen minutes*
The precipitate thus formed was drained off*

This precipitate

contained the sulfate compounds according to the equation,
HgSO^. + NagCOg —***

The resulting gasoline

ilagSO^. + ^ 2 ^

^ ^*^2 *

solution was washed twice and

the water drained off through the separator funnel*
Anhydrous calcium chloride (CaClg) was added to the gasoline
and allowed to stand over night to remove the water*

The

saturated gasoline was then poured off ana found to still
produce a wetting film*
Second Method.

Another method used to remove unsaturat­

ed hydrocarbons from gasoline was to mix with the gasoline an
equal volume of a 5$ solution of sodium carbonate (NagCOg)*
To

this was added a 2 to 5$ solution of pottassium

permanganate

(KMnC^) •

This

mixture was mixed by amechanical

shaker until the purple color changed to brown*

This was

continued until there was no further change in color*
Distilled water was then added and the gasoline thoroughly
washed, after which the water was drawn off by a separator
funnel*

The remaining water was removed by anhydrous

10

calcium chloride

(CaClg) which was allowed to stand in the

gasoline over night*
The liquids given in Table II below were tested for
unsaturation of the hydrocarbons using the bromine test*
Table II
Treated gasoline

Wetting film

Saturated

Treated Kerosene

No wetting film

Saturated

Untreated gasoline

Wetting film

Saturated

Untreated Kerosene

N o we 1 1 i ng film

Saturated In itself
but unsaturated due
to sulfur content

Untreated distillate

No wetting film

Unsaturated

Treated fuel oil
Grade 11A n

No wetting film

Unsaturated

Naphtha

Wetting film

Unsaturated

Alpha Phenyl Beta
pentene

No wetting film

Unsaturated

The main component of turpentine is pinene which is
uns"iturated due to bending and double bond and the strain
of bending is only partly neutralized by the double bond.
It was found that turpentine has a wetting film.

The

structural formula for pinene is given in Fig. 1 Plate I.
Benzene on the other hand is unsaturated due to bending
and double bond.

One effect just neutralizing the other.

C H

H-

CH

CH

H

H

H

V / c

H

CH,

H

F/G /

F IG

C Hi

Z

C H3

C M

Cv
H—

C

Cv

C

H— C

CM

/ \

H— c

C -- H

C— H

„—
H

C

H

C

/ \

C—

H

C -- H

\ y
H

H

CH.

F/G 3
H
I
c

c

I

II

c
H

I
H

C

H

c

-H

H

V /

I

I
=

c ---

C

H

H
FIG

4

H

H

PLATE I

11

Benzene is therefore a very stable saturated compound.

It

was found in this investigation that benzene does not have
a w e t t±ng film.

Pig. 2 Plate I shows the benzene structural

formula•
Xylene like benzene is a perfectly saturated compound.
It is unsaturated due to bending and double bond.

These two

effects exactly neutralise each other forming a very stable
saturated compound.
meta-, and para-,

There are three kinds of xylene, Ortho-,

each having a formula 05114 (0113)2 but a

different structural formulae, Pig. 3 Plate I.

In all

probability the xylene used in this investigation was a
mixture of the three different xylenes.
The compound alpha phenyl beta pentene is unsaturated
and has the structural formula shown in Pig. 4 Plate I.
Alpha phenyl beta pentene was found to have no wetting film.
From the above results it can be seen that the creeping
or wetting film in the case of mixtures of Hydrocarbons
does not depend upon the saturated or unsaturated condition
of the hydrocarbons.
Low Temperature Combustion.

Low temperature combustion

was investigated as another possible source of ohe energy
required to produce wetting films.

12

A jet of oxygen was passed into uhe vapor above liquid
gasoline.

It was found that the gasoline wetting film

established a point of equilibrium about 2 . 5 cm. above the
surface of the gasoline in the container.

After equilibrium

was established gasoline tears rolled down the wetting film.
These tears did no is readily pass into the surface of the
gasoline but seemed to have a tendency to return up the
wetting film.
be formed.

If the oxygen is turned off more tears will

If oxygen is again admitted just as a tear gets

near the gasoline surface it was found, that the tear did
not enter the surface, but entered a new wetting film which
has been formed b y the admission of oxygen. . This new wetting
film crept up the surface of the old wetting film.
Workers in the field of low temperature combustion have
concluded that all surfaces apparently have the power to
produce the effects of confined, flameless combustion when
at h igh temperatures.

Large surfaces will Ignite inflammable

gas mixtures at a lower temperature than small ones.
It is known that ionization does not play a primary
role in flame propagation and that it is manifest rather as
a result than a cause.

Ionization is one of the later steps

of combustion.
Many tests were made to determine if low temperature

13

combustion was a factor in the production of wetting films*
A type K, Leeds and Northrup potentiometer was used in
conjunction with thermo junctions of copper and constantin.
One junction was placed in the liquid gasoline while the
other was so arranged that it could be placed anywhere in­
side the container, either in the wetting film or in the
gasoline vapor.

Another set of junctions was arranged in

such a way that the difference in temperature between the
atmosphere and the liquid gasoline could be determined*
The liquid gasoline was used as the reference temperature*
The container with its thermo junctions and liquid gasoline
was placed in a constant temperature bath*
After equilibrium was established, with the exploring
thermo junction five centimeters above the liquid gasoline,
a small amount of air was admitted*

It was found that the

gasoline vapor was slightly heated*

Upon admitting more air

the exploring junction was cooled but quickly reestablished
the condition of equilibrium.

Measurements indicated that

the gasoline vapor was slightly warmer than the liquid
gasoline•
With the exploring thermo junction placed In the wetting
film ,one centimeter above the liquid it was found upon
admitting air near the top of the container that the junction
in the wetting film was cooled.

A similar result was

14

obtained when the exploring junction was placed in the wet­
ting film two centimeters above the surface of the gasoline.
When oxygen is admitted to a tube containing gasoline
a succession of wetting films rapidly followed one another
notwithstanding the fact that the amount of disturbance
transmitted to the surface is very small.

Illuminating gas

produced more rapid formation of wetting films than oxygen.
The admission of air causes a wetting film to rise rapidly
up the sides of a glass container.
Early Observers of Wetting Films.

James Thomson(9)

in 1855), explained "the curious motions commonly observed
in the film of wine adhering to the inside of a wine-glass,
when the glass, having been partially filled with wine, has
b een shaken so as to wet the inside above the general level
of the surface of the liquid , ’1 by the ’’consideration that
the thin film adhering to the Inside of the glass must
very quickly become more watery than the rest, on account
of the evaporation of the alcohol contained in it being more
rapid than the evaporation of the water,

the watery

portions having more tension than those which are more
alcoholic,

drag the latter briskly away."

Maxwell (10) accepted James Thomson1s explanation and
called attention to the reference to the phenomenon in

15

Proverbs xxiii, 31.
Hall (11), while trying to prove that the angle of
contact of alcohol was zero, which was doubted by many
observers at this time, proposed to check the angle of
contact.

TfIf there is a finite contact-angle for each

plate, the values found for T cos

will be different.

If

these values are all alike, it is reasonably certain that
the contact-angle is zero."
tabular form.
tin, glass.

Hall then gives the results m

He used plates of platinum,

zinc, silver,

To compensate for the observed discrepency

in zinc and silver, Hall says, ” the zinc plate was slightly
wedged shaped and not very regular on the edge.

There was

no such irregularity in the silver plate; but on watching
closely I could see that the alcohol crept rapidly up the
plate, wetting it to a height of several millimeters above
uhe normal line of contact.

Several other trials with a

silver plate showed the same peculiarity.

There may have been

something similar happening on the zinc plate, but if so
I failed to detect it.”
Thomson and Maxwell assumed the solid walls to have
been wet by mechanical shaking and did not observe a wetting
film creeping up a dry suriace.

Hall apparently observed an

alcohol film wetting a silver surface, and believed it

16

accounted for variations in his determinations of the angle
of contact of alcohol with silver and zinc, although he
observed no wetting film on the surface of zinc, mica,
platinum,

or glass*

The single instance of spreading

observed by Hall can be explained on L a n g m u i r 1s theory,
as alcohol contains an OH radical*

It thus appears that

true wetting films have not b e e n given previous study.
Surface Tension*
films and surface

The relationship between wetting

tension was next investigated*

According

to L a p l a c e fs Theory of Surface Tension (12) the contractile
force of liquid surfaces is attributed to the attractions
of the molecules Immediately below the surface of the
liquid for those on the surface, producing a tendency for
the surface molecules to move Into the interior*

Foley (13)

has offered the following objections to Laplace1s theory.
The magnitude of the force normal to the surface would depend
on the curvature of the surface ana would be greater at a
convex surface then at a flat or concave surface*

Consequent­

ly the rise of water in a capillary tube would be due to the
fact that the downward pressure of' the film outside the tube
is greater than the downward pressure of the film insxde
the tube.

Two simple experiments show the lalacy of

L a p l a c e 1s theory.

17

Suppose a long capillary tube be taken and its lower
end placed some distance into the water and the height of*
the capillary column noted*

Now drop some soap solution on

the water outside the tube and thus l^wer the tension out­
side*

If the liquid is supported by the excess of pressure

outside the tube, the height of the capillary column should
be lessened*

On the contrary the height remains constant

for some time*
If the experiment be repeatea but this time placing
the soap solution in the capillary tube, the tension inside
the tube is reduced (demanding a reduced pressure inside)
and the outer pressure remaining constant,

it would seem

according to Laplace that the excess of the outside pressure
would be increased and that the water should rise in the
capillary*

Instead of rising it immediately falls*

Randall, Williams, and Colby (14)
explanation of surface tension.

give a very acceptable

Suppose in a container the

molecules in the surface are considered.

They will have an

attraction for any molecular attraction.

Likewise molecules

within this distance from the walls of tne container will
exert a force of attraction for the wa-lls.

Suppose now that

a molecule is slipped out of the surface layer by some means
or other*

When this happens the surface at that instant is

18

slightly smaller then it was previously.

This diminution of

the area of the free surface causes the phenomenon known as
surface tension.

This dynamic explanation of surface tension

constitutes a more useful conception of the phenomenon.
The surface tensions of naphtha,

turpentine, and

gasoline were measured with a platinum ring on a torsion
du Nouy tensiometer.

The values of the surface tensions

are given in the following table.
Table III
Temperature
Degrees
Centigrade
43
40
38
36
35
34
32
30
28
26
24.5
24
23

Surface Tension in Dynes / Centimeter
Naphtha

25.11
25.18
25.46
25.73
25.80
26.01

Turpentine

Gasoline

29.24
29.24

24.08
24.22
24.63
24.91

29.72
29.86
29.93
30.27
30.62
30.82
30.89

25.04
25.18
25.32
25.46
25.46
25.87

Effect of Saturated Vapor above Gasoline Surfaces.

If a

container partly filled with gasoline and sealed is allowed

19

to stand the space above the gasoline becomes saturated
with, gasoline vapor providing some liquid gasoline remains
in the container*

When this condition of saturation is

reached it is impossible to start a wetting film by means
of a wavelet on the liquid surface#
Me a surement s of Pseudo Visc o s i t y #

Measurements of the

pseudo viscosity were taken for gasoline, naphtha, and
turpentine#

The measurements were taken with a Tag-Saybolt

Universal Thermostatic Viscosimeter•

In this type of

instrument the time in seconds is determined for the flow
of a definite amount (60 cc#) of the material at the required
temperature through a given opening#

The temperatures and

times of flow are given in the following table.
Table IV
Temperature
Degrees
Centigrade
6
10
20

30
40

Time of Plow in S e c o n d s .
Naphtha

Turpentine

Gasoline
29.2

30 #0
30.0
30.0
33 #1

35.6
36.0
34.1
33*0

29.4
29.6
30.0

The pseudo viscosity measurements include all factors
affecting resistance to shear as well as the true viscosity

20

of the liquids;
would,

the latter would Increase, while the former

in general, decrease with increased temperature*

Table IV indicates a negliable change of pseudo viscosity
with change of temperature in the case of the three liquids
examined,

and that wetting films are not caused by the

small changes in temperature due to surface combustion.
Method of M e a s u r e m ent♦

The source of the energy

required to produce wetting films was found by means of
the Universal Interferometer used in the measurement of
the Radius of Molecular Attraction.

It was designed and

constructed in the Barney Laboratory at Denison University.
Plate II shows a plan of the instrument.

A beam of light

from a mercury arc, made monochromatic by the use of filters,
Is incident upon a plate silvered with a 50$ film on the
side toward the source.

In order to utilize tho largest

possible aperature the beam must have an angle of incidence
of 45 degrees.

At the 50$ film the b eam is divided into

transmitted and reflected portions each meeting an end
mirror (3 and 4 ) adjusted so as to reflect the light beams
parallel to each other and to the dividing plate.

Plate II

shows the angles which the end mirrors make with the axis
of the instrument#

The beam transmitted by the plane

parallel dividing plate

jls

parallel to the Incident beam

Weffing Film

& i = w = i
X
>r

3/c/e View
AfercUry A r

c

o
/
/
/
/
/
777,

7/

To/) View
C a fh e + o /n e te r

21

and is reflected by the end mirror (3 ) in a direction
parallel to the dividing plate and the axis of the
interferometer.

The angle 2x, therefore, equals 45°.

The

angle between the end mirror and the axis of the instrument,
or the dividing plate, is 90°-x or 67^°.

The transmitted

and reflected rays are reflected by the end mirrors to a
large mirror m T •

If they meet this mirror at perpendicular

incidence they are returned upon themselves to the dividing
film where they undergo a second division.

The reflected

portion of the transmitted ray and. the transmitted portion
of the reflected ray are received in the telescope of a
cathetometer, and Fizeau Interference fringes oDserved in
the plane of the mirror m 1 •

If the instrument is to be

used as a compound interferometer the mirror m f is tilted
so that the b e a m is reflected to the mirror mg which xs
elevated just enough to allow the beam coming from the end
mirrors to pass under to mirror m » •

The angle between the

mirrors m ! and mg can be adjusted so as to produce any
desired number of reflections between them, the interfering
beams eventually meeting one of the mirrors at perpendicular
incidence and returning upon themselves,

producing

interference bands of any desired order.
The wetting film to be measured was placed on an optical

22

plate placed between m f and mg in such a way that clear
glass was in one of the two interfering paths and glass
with the wetting film was in the other.

The glass plate

extended through both paths so that no correction for the
increased optical path was necessary.
The bed plates carrying the dividing mirror, end
mirrors,

compensating plate, the mirror mg, and the mirror

m f are clamped on a large bronze tube.

They can therefore

be moved to any desirable position on the tube.

The end

mirror (4) is fastened directly to the b e d plate in such a
way that it can be rotated about its center.
(3)

The mirror

can also be rotated about its center and is connected

through an opening in the bed plate to the movable carriage
of a small but exceedingly accurate interferometer.

This

makes it possible to move mirror (3) forward and backward and
in this way lengthen or shorten the optical path of this beam.
Because of the difficulty in producing a collimated
beam of light the adjustment of the mirrors can best be
made using a beam of sunlight.

The successive images are

seen on the mirrors m f and mg, and the order of the
c o m p o u n d i n g can easily be

determined.

The optical plate upon which the wetting film was
allowed to creep was divided into two parts by a strip of

23

collodion*

On the bottom of1 half of this plate was fitted

a container coated with collodion.
The material under observation was placed in the
container and the wetting film formed on half of the optical
plate.

This placed the wetting film in one optical path of

the interferometer and at the same time provided a means of
placing the same thickness of optical glass in both paths*
The formula used in calculating the thickness of the

6X

wetting film is

* where D is the thickness of

N(ni-l)
the wetting film,

$

the interference fringe displacement

in terms of number of fringes, A the wave length of light
employed, N the number of times the light passes through
the wetting film, and n T the index of refraction of the
wetting liquid.

The total change in optical path is

SA ,

S A /N the change in optical path due to one interfering ray
passing once through the

vetting film, n !-l the difference

between the index of refraction of the wetting film and
the index of refraction of air which the wetting film
displaces .
The fringe displacement was measured with a cathetometer
fitted with a micrometer eyepiece.

In every case many

readings were taken and the average of the readings used in
the calculations.

24

The index of refraction of the material was measured
w i t h an Abbe refractometer, and here again the value taken
was the average of many readings•

The following is a table

of the indices of refraction of several liquids as measured
by the Abb& refractometer, the temperature in each case
being 2 2 ° centigrade.
Table V
Saturated Gasoline
Unsaturated Gasoline

1.4209
1.4302 to 1.4279

Naphtha

1.4191

Turpentine

1.4720

Benzene

1.4980

Xylene

1.4953

Ethyl Alcohol

1.3629

Acetone

1.3590

Thickness of Wetting F i l m s .

The thickness of the

wetting films of naphtha are given in Table VI.

The three

groups of results are for the same naphtha but the values
represent longer exposure to the atmosphere.

The naphtha

in group 2 h a d been exposed longer than the naphtha in
group 1.

Likewise the naphtha in group 3 had been exposed

longer then the naphtha of group 2 .

25

Table VI
Group 1
Temperature
Degrees
Centigrade

Thickness
of Wetting
Film in
Centimeters

23.5
24.0

0.0000311
0.0000356
Group 3

Group 2
Temperature
Degrees
Centigrade

Thickness
of Wetting
F i l m In
Centimeters

Temperature
Degrees
Centigrade

Thickness
of Wetting
Film in
Centimeters

18.0
19.0
20.0
21.0

0.0000335
0.0000392
0.0000443
0.0000479

16.0
18.0
19.0
25.0

0.0000427
0.0000478
0.0000494
0.0000536

In making the above measur ments the green line of the
mercury arc o *0000546073 centimeters was used.*

The values

of n* , the index of refraction of the various materials
used, are given in Table V.
Table VI shows that the longer a given liquid is
exposed to the atmosphere the thicker becomes the wetting
film for any given temperature.

There is also an Increase

in the thickness of the wetting film resulting form an
increase in temperature.

26

The diagram, Fig. 1 Plat© IV, shows a typical westing
film of naphtha.

In the center of the diagram can be seen

interference patterns which are irregular in n a t u r e •

This

is due to the debris which has b een carried to the top of
the wetting film and left there when the film drains, and
also to tears which are formed at the top of the film.

The

narrow interference bands in this region show that the tears
and debris are apporximately ten times the thickness of the
wetting films, which is determined by measuring the displace­
ment of the FIzeau bands in the upper and lower portions of
Fig. 1.
U pon observing the wetting films witn a telescope they
were seen to split at the top one or more times.

It was

also observed that new wetting films creep up over the old
wetting films and these new films were extremely fl«.t across
the top and had a very sharp radius of curvature at the upper
edge.

This suggested that wetting films vary in thickness

by some constant.

This was found to be uhe case*

Taking

the values for the thicknesses of the wetting films with
their corresponding temperatures from group 2 of Table VI
and obtaining all of the possible combinations by subtract­
ing the various thicknesses from ea.ch other, one obtains
the thicknesses of film which correspond to the different
degrees of temperature, as shown In Table VII.

27

Table VII

Temperature
Difference
in Degrees
Centigrade
1
1
1
2
2

3

Difference In
Thickness of
Films

0.0000057
0.0000051
0.0000036
0.0000108
0.0000087
0.0000144

Difference in
Thickness of Film
per Degree Change
in Temperature
0.0000057
0.0000051
0.0000036
0.0000054
0.00000435
0.0000048

If the difference In thickness of film per degree
change in temperature are now divided by the radius of :.j
-7
cm. it Is found that they
molecular attraction, 1.5 x 10
are exactly multiples of this value.
accordingly 38.00R,
32.00R.

The results are

34.00R, 24.00R, 36.00R, 29.00R, and

It appears from this calculation that the wetting

films split along planes of cleavage one R apart.

Successive

wetting films differ In thickness by R or multiples of R.
Flow of Liquid in Wetting F i l m s .

Wetting films have

been observed 1 0 cm. above the liquid surface;

the height

depending on the solid surface, the nature, age, and
temperature of the liquid, and the degree of saturation of

28

the vapor space.

A study of the phenomena illustrated in

Pig. 1 Plate IV showed the presence of vigorous motions in
the wetting film.

While under fixed conditions wetting

films always rose to an equilibrium point there were many
indications that this equilibrium was a dynamic one.
Observations with the interferometer showed the formation
of successive wetting films creeping up over the preceding
film.

The motion wich was continuous as long as the vapor

space was unsaturated,

could be explained on the assumption

that the liquid adjacent to the solid was descending while
the outer portion was ascending.

Attempts were made to

observe the oppositely moving currents by means of suspended
particles.

As the thickest wetting films measured were only

one wave length thick it is readily seen that a particle
small enough to be suspended in the downward flowing liquid,
contacting neither the solid wall nor the upward flowing
liquid, would be beyond the limit of resolution of the best
compound microscope.

The continuous upward movements were

readily observed, but the particles were caught by the solid
wall when the liquid current turned downward.
The predicted phenomena was readily observed, however,
when a creeping liquid, containing small suspended particles,
was introduced into a receiver,

the bottom of which was

29

concave upward*

The plano-convex drop of* liquid thus formed

was surrounded by a wetting film*

The microscope plainly

showed the suspended particles rising in the center of the
drop,

flowing outward in the upper protions and toward the

center in the lower portions of the drop.

The vigorous

motions were maintained as long as evaporation continued*
The Effect of Age and Temperature on the Surface
Tension of Naphtha *

Table VIII shows the effect of temperature

on surface tension for fresh naphtha and for stale naphtha;
that is naphtha which was exposed to the atmosphere in
order that the more volatile fractions might evaporate*
Table VIII
Temperature
Degrees
Centigrade
37.0
35.0
32.0
30.0
28.0
26.0
24.0
23.0
22.5

Surface Tension in Dynes / Centimeter
Fresh Naphtha

25.11
25.18
25.46
25.73
25.80
26.01

Stale Naphtha
26.49
26.49
26.69
26.98
27.19
27.59
27.66
28.21
' 28.44

The curves representing these values are given in
Fig* 2 Plate III*
Evaporation at Curved Surfaces•

Surface tension,

30

according to Ewing (15), is the skin of a liquid at any
surface which separates liquid from vapor and is the seat
of a definite amount of potential energy.
results of surface

One of the

tension is to make the conditions of

equilibrium between liquid and vapor depend upon the
curvature of the liquid surface.

The normal conditions for

equilibrium between vapor and liquid apply only to flat
surfaces.

In this condition ohe vapor, at the same tempera­

ture as the liquid, has what is called a saturated, vapor
pressure for that temperature.

In consequence of surface

tension a small drop will evaporate into an atmosphere
that would be saturated or even supersaturated with respect
to large drops or large quantities uf thw liquid, because
the vapor pressure which is required to prevent evaporation
from the curved surface of a small drop is greater than the
vapor pressure which will prevent evaporation from a rlat
surface 01

the same liquid at the same temperature.

The

equilibrium vapor pressure for a small drop is higher than
the normal saturation pressure.
Again as the result of surface tension, the energy
contained in a drop of liquid is greater than the energycontained in an equal volume of liquid when it f^rms a
part of a mass of the same liquid at the same temperature.

31

As, a small drop contains more energy, due to surface tension,
per unit mass than a large drop of "che same liquid at the
same temperature, because the small drop has a relatively
larger area.

The area 01

uhe surface layer can oe taken as

a measure of the stored energy*
W hen the thickness of a vertical liquid film is reduced
to eight times the radius of molecular attraction its
surfaces are parallel*

If such films are stretched, or

allowed to evaporate, they split in the middle,
in stable form 4R thick.

and exist

Such films, known as the first

and second black, because of their low reflection power, are
too thin to produce distructive interference between the
light rays reflected from the front and back surfaces*

Films

of greater thickness than the first black show N e w t o n ’s
interference bands in colors but they store the same poten­
tial energy in their surfaces.
If a drop is evaporating under conditions which keep
its temperature constant,

energy must be supplied in propor­

tion to its loss of mass to provide for its latent heat of
vaporization.

The drop will lose some of its surface energy

because it is decreasing in size, and this supplies part of
the latent heat,

the remainder coming from the outside.

A

drop will evaporate into a space which Is saturated for the

32

same liquid in large quantities.

Thus it is seen that there

can be no equilibrium between a drop and a surrounding atmos­
phere of saturated vapor.

As the drop gets smaller a point

is reached in which the loss of potential energy due to the
contraction of the surface supplies all the latent heat of
vaporization.

After this point is reached the drop will

continue to evaporate without any outside supply of heat.
The equilibrium between a liquid and vapor depends upon
the curvature of the surface between them, if they are at
the same temperature.

Liquid with a flat surface is in

equilibrium with the vapor above it when the vapor is at
the pressure of saturation and there is no tendency on the
whole for the liquid to evaporate or the vapor to condense.
A liquid having a convex curved surface is in equilibrium
with the vapor only when the vapor pressure exceeds the
normal saturation pressure by a definite amount, that is,
only when the vapor is sufficiently supersaturated.
The degree of supersaturation necessary for equilibrium
depends upon the curvature of the surface, in a manner first
established by Kelvin (16).
lop^P1 _
~

2S

K e l v i n ’s formula is

where P ! is the equilibrium vapor pressure

RT/Or

over a curved surface and P s is normal pressure of satura­
tion, S is the surface tension in dynes per linear centi-

33

meter*

By assuming a perfect gas RT was put equal to P V •

Tile value of R T is for a gram molecule, and must be divided
by the molecular weight of the liquid to reduce the value
to cgs units, y o

is the density of the liquid and r the

radius of a drop of liquid assumed to be spherical*

This

formula determines the pressure P* which must be maintained
in the supersaturated vapor around the drop in relation to
the normal pressure of saturation P g for the same temperature,
if the drop just escapes shrinking by evaporation*

An

increase in p» for which the value has been calculated would
cause the drop to grow*

A drop smaller then that for which

the value was calculated would disappear by evaporation;
likewise a drop having a value of r larger would grow*
It is only when the drop is exceedingly small that the
excess of P* over P s is considerable.

Kelvin calculated

that for water vapor at 10°C, RT (which is treated as equal
q
to PV) is 1*30 x 10 cgs units# The surface tension uf
water at that temperature above 76 dynes per linear
centimeter, a n d ^
log-in P* —
pj

is one gram per cubic centimeter* Then

1.01 9 where D is the diameter of the drop in
d ~

m i H i o n t h s of a millimeter.

The formula gives for drops of

water the values given In the Table IX*
If K e l v i n 1s formula is applied to naphtha drops at

34

10° centigrade it becomes logn/%
P* _ 1.2066,
where D is the
10 ^
—
diameter of the drop of naphtha in millionths of a millimeter.
When Kelvin*s formula is rewritten to apply to the case
of a thin vertical film, such as that produced in the capillary
space between two parallel plates, it becomes logA P* —
2S
0 P s ~ RT /o (2d)
where d is the distance between the parallel plates.
If d is given in millionths of a millimeter, we have in
<-*
9
the case of naphtha at 10° C, that RT ^ 0.273 x 10 cgs units*
The surface tension of naphtha is S s 28.3456 and its density

/& = 0.7466.

Therefore log.,^ P* ^
t i

0.6033.
a-

This formula

expresses how large the pressure P* must be In the
supersaturated vapor around the top of the wetting film,
which has a radius of curvature d, in relation to the normal
pressure of saturation P g , for the same temperature, on the
flat surface of the wetting film.
The ratio P*/Pa for different values of d given in
Table IX below for naphtha are plotted along with Kelvin* s
value for water vapor In Fig. 1 Plate III.

35

Table IX
Diameter of
Drop of Water
in millionths
of a millimeter

Ratio of P*/Pa
for Water Vapor

100

1.02

50

1.05
1.26
1.59
3.2

10

5
2
1

10.2

Diameter of
Ratio of P f/P_
Drop of Naphtha
in Millionths
for Naphtha
of a Millimeter
Vapor

1.014
1.028
1.149
1.320
1.415
1.589
2.003

100

50
10

5
4
3
2

Radius of
Curvature of
Wetting film of
Naphtha against
a Single Plate
in Millionths
of Millimeter

Ratio of P'/Ps
for Naphtha
Vapor at Top <
Wetting F ilm

100

50
10

5
4
3
2

1.5
4.012

1

1

1.028
1.060
1.320
1.743
2.003
2.525
4.012
6.370
16.094

For large radii of curvature the ratio P !/ P s for the
three curves approaches unity, because then the surface is
becoming flat,

Chamberlain’s value for the radius of

molecular attraction, 1.5 x 10"7 cm., is now generally
accepted*

Chamberlain also determined the minimum tnickness

to

Cj

W A T E R /N

C A R /l-L A R Y

TU B E

N A P H T H A IN CAR/LL A R Y 'TUBE

f-

n a p h th a

on

1
3

4

v e r t ic a l .

1

pla te

i
to

5

op A

P ftL L fO N T H S

PitE E tP7E T E R

RIG t

37

sta le

^ 30

naphtha

R R E S H NA PHTHA

25

22

27

25
SURPACE

TENSION

R /G

/N

2.

28

29

D Y A /E ^ /c p f

PL A T E m

36

of a stable liquid film to be 4R, or 6 millionths of a
millimeter.

At this value the curves begin to separate

rapidly and turn upwara.

W hen the radius of curvature

equals the radius of molecular attraction, that is at
1*5 millionths of a millimeter, P T/P

s

is large, and for a

value slightly less than this a drop would evaporate in an
explosive m a n n e r •

Surface tension exists in a film 1*5

milljLonths of a millimeter in thickness.

If it were possible

to form a film of less than this thickness its surface tension
would be weakened.
vapor,

In other words it would explode and become

since its energy content is equal to the latent heat of

vaporization.

That this is the case is shown by the rapid

increase in the ratio of P^/Pg in the region of 1.5 to 1
millionths of a millimeter.
Notwithstanding the large differences in surface tension
and density between water and naphtha to the curves in Pig. 1
Plate III show similar characteristics*
A steel straight edge 11 cm. h n g

and U .u2 cm. thick,

the upper edge of which had been sharpened to a razor edge,
was placed in on« path and parallel to the axis of the
interferometer.

The lower edge was submerged m

naphtha.

IJM

riffi

FIG Z

P L A T E IF

37

The appearance of the FIzeau hands is shown in Fig# 2
Plate IV.

The curvature of the Fizeau bands is due to the

evaporation of naphtha.

The Increasing distance between

bands from bo t t o m to top Is due to the vapor density
gradient which existed above the naphtha surface.

The

curvature of the bands from the flat side of the straight
edge outward indicates the evaporation from the flat surface
of the wetting film*

The band just above the straight edge

shows a kink, due to the high rate of evaporation from this
sharp edge. . The fact that the band is curved at all indicates
vary rapid evaporation, because the area of liquid at this
curved surface is very small and a large amount of material
would have to evaporate to show this curvature in a band
above the sharp e d g e •
Summary.

1.

Wetting films are formed by liquids

containing two or more components of different vapor tensions*
2.

Materials which produce no wetting film are found to

be able to produce a wetting film when mixed if the liquids
have different vapor tensions.
3*

Saturation and unsaturation of the hydrocarbons

play no part in the production of wetting films.
4.

Low temperature combustion Is not sufficient to

contribute appreciably to the production of wetting films.

38

5.

The variation of pseudo viscosity with temperature

does not contribute appreciably to the production of
we 11 ing fi1ms •
6.

The thickness of wetting films increases with rise

in temperature; due to the increase in the true viscosity
of the liquid mixture•
7.

The thickness of wetting films increases with the

age of the liquid mixture*
8*

Wetting films have planes of cleavage which are

exact multiples of the radius of molecular attraction.
9.

Differences in surface tension due to localized

evaporization play a primary role in the lifting of a
wetting film.
10.

The small radius of curvature at the top of a wet­

ting film, resulting in a high rate of evaporation producing
a region of high surface tension and supplies the energy
necessary to move the superficial layer upward.
with it some of the body of the liquid.

This drags

The thickness, of

the upward moving layers, depends on the true viscosity of
the liquid mixture.

Debris collects at the top of a liquid

film in much the same manner as drift material on a beach.

39

References♦

(1)* Rolt, F.H* "Gauges and Fine

Measurements” Macmillan and Co. Vol. 1 p 72 and 168 (1929).

(2).

Chamberlain, C.W. "Radius of Molecular Attraction”

Rhys. Rev. Vol. 31, p 170 (1910).
(3).

Quincke. Phil. Mag. Vol. 38 p 95 (Aug. 1869).

(4).

Reinhold and Rucker. Phil. Trans. Vol. 172 p 447

Part II (1881).
(5).

Reinhold and Rucker. Phil. Trans. Vol. 174 p 645

Part II (1883).
(6).

Reinhold and Rucker. Phil. Trans. Vol. 184 p 505

(1893A).
(7).

Johonott and Chamberlain* Phil. Mag. Vol. 47

p 501-522 (June 1899).
(8).

Langmuir, I. National Academy of Science Vol. 3

p 251 (1917).
(9).

Thomson, James. Phil. Mag. Vol. 10 p 330 (1855).

(10).

Maxwell. "Theory of Heat” Longmans, Green and Co.

p 300 (1921).
(11).

Hall, T.P. Phil. Mag. S5 Vol. 36 p 400 (Nov. 1893).

(12).

Laplace. Phil. Mag. Vol. 16 p 315 (Oct. 1883).

Lord Rayleigh* s discussion of Laplace. Laplace. Mecanique
Celeste, Supplement au Livre X.
( 13).

Foley, A.L. (Abstract)

Theory of Surface Tension.”

"Objections to Laplace*s

Proceedings of the Indiana

Academy of Science p 181 (1910).

40

(14).

Randall, Williams, and Colby. "General College

Physics" Harpers and Brothers Publishers p 142-5 (1929).
(15).

Ewing,

Sir J.A. "influence of Surface Tension on

Condensation and Ebullition."

Dictionary of Applied Physics

Macmillan and Co. Vol. I p 962-6.
(16).

Kelvin. Phil. Mag.

Bibliography.

(1922).

(4). Vol. 42 p 448 (1871).

Bravais. Ann. Chem. Phys. V P 492.

Buys - Ballot. Progg. Ann. lxxi p 191 (1847).
Bone and Townsend.
Longmans,

"Flame and Combustion In Gases"

Green and Co. p 33.

Bone* j . Roy. Soc. Arts 62, 787, 801. 818 (1914).
Bennett and Mardles. J. Chem. Soc. 3155,

(1927); 870

(1928).
Devaux, H. Washington,

Smithsonian Inst., Ann. Rep.

261-273 (1913).
Donncn, F.G. and Barker, J.T. London, P r o c . R. Soc.
(A) 85, 557-573 (1911).
Duprfe. Theorie M&canique de la Chaleur,

(1869).

Foley, A*L. Proceedings of the Indiana Academy of
Science p 67 (1895).
Foley, A.L. Phy. Rev. Vol. 3 No. 5, p 381 (1896).
Faraday.

"Experimental Researches on Electricity ’1

Phil. Trans. Vol. cxxiv (I) p 55 (1834).
Gauss. Principia Generalia Theoriae Figurae Fluidorums
in Statu Aequilibrii.

41

Hardy, W-B. London, Proc. R. Soc.

(A) 8 6 p 610-636

(1912).
Hagen. Progg. Ann. lxxvii p 453, 464, 449, 456 (1849).
Hagen. Progg. Ann. lxvii pp 1, 153 (1846).
Hoover and Redi. J. Am. Chem. Soc. 43, p 882 (1921).
Langmuir,

I.J. Amer. Chem. Soc. Easton Pa. 38 p 2221

(1916).
Lenard. Wied. Ann. xxx p 209 (1887).
Langmuir. J. Am. Chem. Soc. 42, 2190-2205 (1920).
Langmuir. J. Am. Chem. Soc. 38, 1145, 2221 (1916).
Marcelin, A. Paris, Ann. Physique 1, 19-34 (1914)•
Milner,

S.R. Pil. Mag. London 13, p 96-110 (1907).

Magie. Wied. Ann. xxv p 432 437 (1885).
Magie. Phil. Mag. p 171 (Aug. 1888).
Merian. Pogg Ann. lxxiii p 485 (1848).
Marek - Hahn. !lCatalytic Oxidation of Organic
Compounds in the Vapor Phase.” Amer. Chem. Soc. Monograph
Serirs No. 61.
roisson. Nouvelle Theorie de L !Action Capillaire:
Paris

(1831).
Pockels, A. Nature, London 43 p 437-439 (1891).
Quincke. Phil. Mag. Vol. 41 p 245, 370, 454 (1871).
Quincke. Pogg. Ann. cxxxv p 641 (1868).

42

Quincke, Wied. Ann. xxvii p 219 (1886).
Quincke. Pogg. Ann. cxxxix p 1 (1870).
Rayleigh. Phil.

Mag. Vol.

16 p 309 (1883).

Rayleigh. Phil.

Mag. London 48, p 331-337 (1899).

Rayleigh. P r o e .

Roy. S o c . xxix p 71 (1879).

Rayleigh. Phil.

Mag. Vol.

xxx p 285, 386, 456 (Oct. 1890) .

Riley, J.T. Phil. Mag. Vol. 15 p 191 (Feb. 1883).
Szyszkowski, B. v., Zs, physik Chem., Leipzig 64,
p 385-414 (1908).
Sondhauss.

Pogg. Ann. Erg. viii p 266 (1878).

Sent!s. Journ. de Phys. IX p 384 (1890); or Phil. Mag.
p 564 (Dec. 1891).
Schnabel. Ber. 46, 5 (1913).
Timberg. Wied. Ann. xxx p 545 (1837).
Terquem. Journ. de Physique vii p 406 (1878).
Taylor. "Chemical Reactions at Surfaces *n Chem. Rev.
9, p 1-46 (1931).
Volkmann. Wied. Ann. xi p 177 (1880).
Voille. Journ. de Physique iv p 313 (1857).
Worthington. Phil. Mag. Vol. xx p 51 (July 1885).
V a n der M e n s b r u g g h e . Phil. Mag. Vol. xxxiii p 270
(Apr. 1867).
Werthelm. Pogg. Ann. cii p 595 (1857).
Weinberg. Beibl. xvi p 496 (1892); Journ. de Physique
p 378 (Sept. 1892).

43

Worthington* Proc*

Roy* Soc* xxxii p 362 (1881);

Phil* Mag* Vol* 20 p 46 (Jan* 1885).
Wilhelmy. Pogg* Ann* cxix p 186 (1863)*
Waterston. Phil* Mag. Vol. xv p 4 (Jan* 1858).
Wendt and Grimm. Ind. Eng. Chem. 16 p 890 (1924)*